JP3260587B2 - Time series data management method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a time-series data management system, and more particularly to a time-series data system which is effective for statistical work and the like based on contents of a relational database at an arbitrary time in an online information system having a relational database as a core. The present invention relates to a time-series data management method having a function of retaining data (contents of a relational database at arbitrary plural times) and a function of searching.

[0002]

2. Description of the Related Art Conventionally, in statistical processing relating to the contents of a relational database, it is necessary to collect and manage time-series data as a source at each point in time, and to collect time-series data in order to avoid data inconsistency. This has been done by temporarily stopping the update of the relational database and saving the data to another file.

[0003]

In the above-described conventional method, it is necessary to collect time-series data required for statistical processing at each point in a relational database. In addition, there is a problem that management becomes complicated.

[0004] Further, since it is necessary to stop the update work to relational data in order to collect time series data, it is impossible to perform statistical work in a continuous non-stop operation environment of an online information system. There was a problem.

SUMMARY OF THE INVENTION In view of the above, it is an object of the present invention to continuously access a relational database in order to achieve continuous nonstop operation of an online information system centered on a relational database and to simplify the operation of the relational database. Another object of the present invention is to provide a time-series data management system that automatically collects time-series data while enabling easy access to the time-series data.

As a conventional technique, in a database management method for processing a search using an inverted file, the use of an additional position pointer and an extraction position pointer eliminates the need to manage time-series data on the application program side. A database management method is known (see JP-A-4-354036). However, this database management method does not allow access to time-series data at an arbitrary time point as in the present invention.

[0007]

Time series data management method of the present invention In order to achieve the above object, according to a start point and the end point and the user data is defined as a column, a data table having a table structure, present time
Start point of user data of point and maximum value as initial value
And the latest data that is the user data.
A latest data table storing the latest data table, and a data table having the same table structure as the latest data table.
A relational database containing a time-series data table that holds the pre-update data as time-series data, and the latest data
For judging whether to retain data in the time-series data table
When there is a request for updating the latest data that sets the holding time that is the time when the latest data is started before the holding time, the latest data is used as time-series data and the time-series data table is used. determines that it is necessary to hold the, otherwise, the determining held determination means that it is not necessary to possess the time-series data table the latest data as time-series data, the holdings When it is determined that the holding in the time-series data table is necessary by the determination unit, the latest data is set to the holding time at the end time, and the time-series data is stored in the time-series data table as time-series data together with the start time. set a series data retention means when held by the holders when the latest data when held is performed to the time series data table to its starting point by the time-series data retention means And updated in accordance with the update request on the, other hand, if held to the time-series data table is determined to be unnecessary by the possession determination means, of the latest data
The latest data updating means for updating only the data in accordance with the update request, and when there is a data search request that sets a search target time,
If the search target time is the initial value, it is determined that the search for the latest data has been instructed.If the search target time is other than the initial value, it is determined that the search for valid data has been instructed at the search target time. Searching time determining means for determining, a latest data searching means for performing a search for the latest data table in accordance with the search request when the searching of the latest data is instructed by the searching time determining means; If the search for valid data in the search target point is determined to be instructed by the determination means, the latest data tables and the later <br/> end earlier than the start time from the time series data table A time-series data extracting means for extracting data including a search target time point to generate a time-series data set, and performing a search on the time-series data set in accordance with the search request And a series data retrieval means.

[0008]

According to the time series data management method of the present invention, possession judgment is performed.
Whether the instrument retains the latest data in the time series data table
The holding time, which is the reference time for determining
When there is a request to update new data, at the start of the latest data
If the point is before the holding time, the latest data
Must be stored in time series data table as series data
Is determined, otherwise, the latest data
Can be stored in the time-series data table as time-series data.
Is determined to be unnecessary, and the time-series data holding
Has been determined to be necessary for time-series data
In this case, the latest data table will be
Set as time series data with the start time
In the series data table, the latest data updating means
Data is retained in the time-series data table by the data retention means.
Then, the latest data is set as the holding time at the start time.
And then updates it in accordance with the update request.
Is determined to be unnecessary in the time series data table
Then, update only the latest data according to the update request and search
The point-of-time judging means requires a data search
Request, if the search target time is the initial value, the latest data
Is determined to have been instructed, and the search target
Search of valid data at the time of the search target point otherwise the value
Has been instructed, and the latest data search means
The search for the latest data is instructed by the search time determination means.
If it is determined that the search
Search according to the request, and the time-series data
The point determination means detects valid data at the
When it is determined that the search is instructed, the latest data table
And from the time series data table
Extract data that includes the search target time before
Data set, and the time-series data search means
The data set is searched according to the search request .

[0009]

Next, the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing a configuration of a time-series data management system according to one embodiment of the present invention. The time-series data management method according to the present embodiment includes a latest data update request 1,
Holding time 2, holding determining means 3, time-series data holding means 4, latest data updating means 5, relational database 6, data search request 9, latest data searching means 10, search time determining means 11, , A time-series data extraction unit 12, a time-series data set 13, a time-series data search unit 14, and a search target time point 15. The relational database 6 has the latest data table 7
And a time series data table 8. Holding date 2 is the latest data
This is an arbitrary time point that serves as a criterion for judging whether or not the data is stored in the time-series data table, and is set by the user at the time of the latest data update request.

Referring to FIG. 2, the latest data table 7 and the time series data table 8 have a start time column 21 and an end time column 2 respectively.
2 and a user data string 23 are defined.

Referring to FIG. 3, the processing for updating the latest data in the time-series data management method according to the present embodiment includes a start time / holding time comparison step S101, a pre-update image acquisition step S102, a time-series data table insertion. Step S10
3, a latest data table update step S104, and a latest data table update step S105.

Referring to FIG. 4, the processing at the time of data search in the time-series data management method of this embodiment includes a search target time / initial value comparison step S201, a time-series data set creation step S202, a time-series data set Search step S2
03 and the latest data table search step S204.

Next, the operation of the time series data management system according to the present embodiment configured as described above will be described.

(1) First, the operation when the data update of the latest data table 7 is instructed by the latest data update request 1 will be described.

When the user instructs the latest data update request 1, the user sets an arbitrary time after the start time of the latest data as the holding time 2 so that the time series data is held when the latest data is updated. To instruct.

When there is the latest data update request 1, the possession judging means 3 compares the start time of the latest data to be updated in the latest data table 7 with the possession time 2 (step S101).
If the start time is before the holding time 2, it is determined that the holding of the time-series data is necessary. In other cases, the holding determination unit 3 determines that the holding time 2 is outside the validity period of the latest data, or that the time-series data at the holding time 2 has already been held.

When the possession determining means 3 determines that the possession of the time series data is necessary, the time series data retaining means 4
The pre-update image of the latest data to be updated is acquired from the latest data table 7 (step S102), and the possession time point 2 is set as the end point and inserted into the time-series data table 8 as time-series data (step S103).

Further, the latest data updating means 5 sets the holding time 2 as the start time of the latest data to be updated, and updates the corresponding data in the latest data table 7 (step S10).
4).

On the other hand, when the possession judging means 3 judges that the holding of the time-series data is not necessary, the latest data updating means 5 updates the corresponding data in the latest data table 7 (step S105).

(2) Next, the operation when the data search request 9 instructs data search of the relational database 6 will be described.

When the user designates the data search request 9, the user designates an arbitrary time as the search target time 15, thereby instructing to search for time-series data valid at that time. Alternatively, by setting an initial value at the search target time point 15, an instruction is given to search for the latest data.

When the data search request 9 is received, the search time determination means 11 refers to the search target time 15 and, if the value is an initial value, determines that the search for the latest data is instructed (step S201). Yes), latest data search means 1
0 is searched for the latest data from the latest data table 7 (step S204).

On the other hand, if the search target time point 15 is other than the initial value, the search time point determination means 11 determines that the search for the time-series data is instructed (No in step S201),
Only the time-series data that was valid at the time indicated by the search target time 15 is extracted by the time-series data extraction means 12. That is, the time-series data extraction means 12 updates the latest data table 7
And the start time from the time series data table 8 is the search target time 1
Data that is before 5 and whose end time is after the search target time 15 is extracted, and a time-series data set 13 is created (step S202).

Next, the time-series data search means 14 performs a search process using the time-series data set 13 in accordance with the instruction of the data search request 9 (step S203).

A specific example will be described in more detail with reference to FIGS.

FIG. 5 shows time point 3 serving as a reference for statistical processing and the like.
2 is a diagram showing an example of the relationship between the holding point 2 to be set at the time of updating the latest data and the latest data update timing 33 according to the instruction of the latest data update request 1. FIG.

FIG. 6 shows the contents of the relational database 6 as a result of updating the latest data table 7 at the timing of FIG. 5 corresponding to the latest data update request 1 and the latest data or time corresponding to the data search request 9. FIG. 9 is a diagram illustrating an example of data contents to be searched when searching for series data. The value H in the end time column 22 of the latest data table 7
V is the maximum value (HIGH-
VALUE) is set.

(1) First, the operation at the time of updating the latest data will be described.

As shown in FIG. 5, for example, when time-series data at the end of each month is held , the user is
During the period from time to immediately before 0:00 on August 1, 1994, the value of 9406 was held at holding time 2 and 9
In the period immediately before 0:00 on the first day of the month, a value of 9407 is set at the holding point 2.

In FIG. 6, when the latest data update request 1 instructs to update the latest data 45 at the timing shown in FIG.
With the start time in Now, the starting point in the latest data 45 is 9407, and the holding point 2 indicates 9409
Since it is confirmed that the time-series data at the time is not held, the time-series data holding unit 4 acquires the pre-update image of the latest data 45 from the latest data table 7 and, at the end time, the value 9409 of the holding time 2. Is set, and the time-series data 51 is set.
In the time series data table 8. Next, the latest data updating means 5 sets the holding time 2 at the start time of the latest data 45.
Of the latest data 4 of the latest data table 7 by setting the value 9409 of
Update 5

Next, when the latest data update request 1 instructs to update the latest data 45 at the timing shown in FIG. 5, the possession judging means 3 compares the possession time 2 with the start time in the latest data 45. , In this case, the latest data 4
Since the start time in 5 indicates the time 9409, which is the same as the holding time 2, the time-series data holding means 4 does not hold the time-series data, and only the latest data updating means 5 updates the latest data 45.

(2) Next, the operation at the time of data search will be described.

First, it is assumed that the user designates a data search request 9 by setting an initial value 0 at a search target time point 15 in order to search for the latest data. Since the initial value 0 is set with reference to the search target time point 15, the search time point determination means 11 determines that the search for the latest data has been instructed. At this time, the latest data search means 10 performs a search process on the latest data table 7 in accordance with the data search request 9.

Next, in order to search the time series data as of the end of July 1994, the user sets the search target time point 15 to 94.
It is assumed that 07 is set and the data search request 9 is instructed. The search time determination means 11 refers to the search target time 15 and
It is determined that the search for the time-series data at time 407 has been instructed. At this time, among the data in the latest data table 7 and the time-series data table 8, the time-series data extraction unit 12 determines that the start time is smaller than the value 9407 at the search target time and the end time is not less than the value 9407 at the search target time. Latest data 46
And 47 and time series data 48 and 49 are extracted,
A time series data set 13 is created. Further, the time-series data search unit 14 performs a search process on the time-series data set 13 in accordance with the data search request 9.

[0036]

As described above, according to the present invention, the latest data table for storing the latest data and the time series data table for storing the time series data are prepared in the relational database, and the latest data table and the time series data table are prepared. The start time point column and the end time point column are provided within the, and the possession determining means, the time series data retaining means, the latest data updating means, the retrieval time judging means, the latest data retrieving means, the time series data extracting means and the time series data retrieving means are provided. With the provision, it is possible to simplify the management and operation of the time-series data and the statistical work using the same while maintaining the conventional access performance and user interface for the latest data.

In addition, time series data can be automatically collected when the latest data is updated. In an online information system having a relational database as a core, the relational database update operation and the relational database can be used in a continuous nonstop operation environment. There is an effect that it is possible to simultaneously execute the statistical work that has been performed.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a time-series data management system according to one embodiment of the present invention.

FIG. 2 is a diagram showing a specific example of a data structure of a latest data table and a time-series data table in FIG.

FIG. 3 is a flowchart illustrating processing at the time of updating the latest data in the time-series data management method according to the present embodiment.

FIG. 4 is a flowchart illustrating a process at the time of data search in the time-series data management method according to the present embodiment.

FIG. 5 is a diagram illustrating an example of a relationship between a possession point and a latest data update timing in FIG. 1;

FIG. 6 is a diagram illustrating a specific example of data contents of a relational database when the latest data update and data search are performed by the time-series data management method according to the present embodiment.

[Explanation of symbols]

1 Latest data update request 2 Holding time 3 Holding judgment means 4 Time series data holding means 5 Latest data updating means 6 Relational database 7 Latest data table 8 Time series data table 9 Data search request 10 Latest data search means 11 Search time judgment means 12 Time series data extraction means 13 Time series data set 14 Time series data search means 15 Search target time 21 Start time sequence 22 End time sequence 23 User data sequence 32 Reference time 33 Latest data update timing 45, 46, 47 Latest data 48 , 49,51 Time series data

Continuation of the front page (58) Fields investigated (Int.Cl. ⁷ , DB name) G11B 17/30 Japanese patent literature Foreign patent literature Non-patent literature

Claims (1)

(57) [Claims] 1. A data table having a table structure in which a start time, an end time, and user data are defined as columns, and
The starting point of the current user data and the initial value
The end point when the large value is set and the maximum
The latest data table which stores the new data, a data table having the latest data table the same table structure, the latest data
A relational database that includes a time-series data table that holds the data before update of the data table as time-series data, and determines whether to keep the latest data in the time-series data table
If there is a request to update the latest data that sets the holding time, which is the reference time, if the start time of the latest data is earlier than the holding time, the latest data is used as the time-series data and It is determined that it is necessary to hold in the data table, otherwise, holding determination means that determines that it is not necessary to hold the latest data as time-series data in the time-series data table, If it is determined by the holding determination means that the holding in the time-series data table is necessary, the latest data is set to the holding time at the end time and the time-series data is set as time-series data together with the start time. Time-series data holding means to hold the table, and when the time-series data holding means holds the data in the time-series data table, the latest data is stored at the start time. After setting the points, update according to the update request,
On the other hand, when the holding determination unit determines that the holding in the time-series data table is unnecessary, the latest data updating unit that updates only the latest data in accordance with the update request; and a data search request that sets a search target time. When the search target time is the initial value, it is determined that the search for the latest data is instructed. When the search target time is other than the initial value, the search for valid data at the search target time is instructed. Search time determination means for determining that the latest data table has been instructed by the search time determination means, and a latest data search means for performing a search in accordance with the search request for the latest data table , when the search for valid data in the search target point is determined to be instructed by the searching time determining means, the latest data table and the time-based The starting point from the data table
A time-series data extraction unit that extracts data including the search target time later and before an end time to generate a time-series data set; and a time-series data search that searches the time-series data set according to the search request. Means for managing time-series data.